Skip to content

Dietary therapy

TL;DR — Diet is not an adjunct in IBS; in the only head-to-head trial against optimised drug treatment it beat it. In CARIBS (n=294, Rome IV, moderate-to-severe), 4-week response (IBS-SSS drop ≥50) was 76% for low-FODMAP plus traditional dietary advice, 71% for a fibre-optimised low-carbohydrate diet and 58% for optimised medical treatment selected by predominant symptom (p=0.023) (Nybacka 2024, PMID 38643782). The network meta-analysis of 28 trials and 2,338 patients ranks a starch- and sucrose-reduced diet first for global symptoms (RR of not improving 0.41, 0.26–0.67; two trials) and low-FODMAP fourth (0.51, 0.37–0.70; 24 trials), with low-FODMAP the only diet superior to habitual diet for bloating (0.55, 0.37–0.80); confidence was low or very low for almost every comparison except low-FODMAP and starch/sucrose-reduced versus habitual diet, which were moderate (Cuffe 2025, PMID 40258374). Against traditional dietary advice rather than habitual diet, low-FODMAP has repeatedly failed to show superiority (50% vs 46% responders, p=0.72; Böhn 2015, PMID 26255043). Blinded FODMAP reintroduction shows the restriction can be substantially unwound: 85% of powders triggered recurrence but at an average of only 2.5 ± 2 FODMAPs per patient, with fructans (56%) and mannitol (54%) commonest (Van den Houte 2024, PMID 38401741). Fibre works modestly and psyllium specifically: 52% vs 44% clinical response (RR 1.21, 1.03–1.41), psyllium RR 1.53 (1.06–2.19) across 30 RCTs (Staudacher 2026, PMID 42600900). The unresolved costs are microbial and behavioural: low-FODMAP lowers faecal Bifidobacterium, Actinobacteria and butyrate, so "strict long-term use should not be advised" (Wilson 2020, PMID 32433273).

The head-to-head trial that changed the framing

CARIBS (NCT02970591; Sahlgrenska, Gothenburg; recruitment Jan 2017–Sept 2021) randomised 304 adults with Rome IV IBS and IBS-SSS ≥175 to three arms for 4 weeks, single-blind to diet name, open-label for drugs (Nybacka 2024, PMID 38643782):

Arm n (mITT) Responders (IBS-SSS drop ≥50) Completion
Low FODMAP + traditional (NICE) dietary advice ("LFTD") 96 73 (76%) 91/96 (95%)
Fibre-optimised low-carbohydrate, high-protein/fat diet 97 69 (71%) 92/97 (95%)
Optimised medical treatment by predominant symptom 101 59 (58%) 91/101 (90%)

p=0.023 across groups. Five of 91 in the medical arm stopped early for side effects; two per group discontinued for adverse events in the diet arms; no serious adverse events. The authors' conclusion — "dietary interventions might be considered as an initial treatment" — is the empirical basis for treating diet as a first-line therapy rather than as background advice. Limitations to hold onto: single centre, 4-week primary endpoint, 82% women, mean age 38, and the drug arm was open-label while diets were masked only as to name.

DOMINO and CARIBS together are the empirical case for diet-first. They were run in different settings (Belgian primary care versus a Swedish specialist clinic), used different comparators (an antispasmodic versus optimised symptom-directed medical treatment), used different delivery (a smartphone app versus dietitian-led counselling), and both favoured diet. DOMINO's authors state it directly: "A FODMAP-lowering diet should be considered the first-line treatment for IBS in primary care" (Carbone 2022, PMID 35483886). Post-hoc analyses of DOMINO have examined subtype-specific effects (Di Rosa 2024, PMID 39086991), inflammatory biomarkers (Tack 2025, PMID 41436948) and eosinophil-driven immune activation and its relation to diet response (Routhiaux 2026, PMID 41670574) — the last connecting dietary response to the mucosal immune mechanism described below.

Comparative efficacy across diets

The 2025 network meta-analysis (28 RCTs, 2,338 patients, 11 dietary interventions vs 4 controls; Cuffe 2025, PMID 40258374). Values are relative risks of the outcome not improving versus habitual diet — lower is better.

Diet Global symptoms RR (95% CI) P-score Trials Abdominal pain RR Bloating RR
Starch- and sucrose-reduced 0.41 (0.26–0.67) — ranked 1st 0.84 2 0.54 (0.33–0.90), ranked 2nd not superior
Low FODMAP 0.51 (0.37–0.70) — ranked 4th 0.71 24 0.61 (0.42–0.89), ranked 5th 0.55 (0.37–0.80) — only diet superior to habitual
BDA/NICE dietary advice 0.62 (0.43–0.90) — ranked 10th 0.44 8 — —

For bowel habit (23 trials) no diet beat any control, though low-FODMAP beat BDA/NICE advice (RR 0.79, 0.63–0.99). All network comparisons were low or very low confidence except low-FODMAP-vs-habitual and starch/sucrose-reduced-vs-habitual (moderate). Two points follow. First, the top-ranked diet rests on two trials and should be read as promising, not established. Second, low-FODMAP's evidence base is an order of magnitude larger than any competitor's, which is why it dominates guidelines (guidelines) despite ranking fourth.

Low FODMAP: the trial record

Study Design Result
Halmos 2014, PMID 24076059 (ACTRN12612001185853) 30 IBS + 8 controls, randomised single-blind crossover, 21 days of provided food, <0.5 g FODMAP/meal Overall GI symptom score 22.8 mm (16.7–28.8) on low FODMAP vs 44.9 mm (36.6–53.1) on typical Australian diet (p<0.001); bloating, pain and flatus reduced; controls unaffected
Böhn 2015, PMID 26255043 (NCT02107625) 75 Rome III patients, multicentre, parallel, single-blind, 4 weeks: low FODMAP vs traditional IBS dietary advice Both improved (p<0.0001 within groups); no difference between groups (p=0.62); ≥50-point IBS-SSS response 19/38 (50%) vs 17/37 (46%), p=0.72
Altobelli 2017, PMID 28846594 Meta-analysis of RCTs and cohorts Significant reduction in pain and bloating vs traditional and vs high-FODMAP diets; no significant difference in stool consistency; superiority over conventional IBS diets, especially long-term, "remains to be demonstrated"
Van den Houte 2024, PMID 38401741 (NCT04373304) 117 tertiary-care patients; 6-week elimination then 9-week blinded randomised reintroduction with 6 FODMAP powders vs glucose control IBS-SSS 301±97 → 150±116 (p<0.0001), 80% responders. Recurrence triggered by 85% of powders, mean 2.5 ± 2 FODMAPs per patient. Triggers: fructans 56%, mannitol 54%, GOS 35%, lactose 28%, fructose 27%, sorbitol 23%, glucose control 26%
Nybacka 2024, PMID 38643782 CARIBS (above) 76% response, beating optimised drug therapy
Tunali 2024, PMID 38717025 121 patients randomised to microbiome-based AI-personalised diet (n=70) vs low FODMAP (n=51), 6 weeks IBS-SSS change −112.7 vs −99.9 (p=0.29) — no significant difference; both improved significantly on IBS-SSS, anxiety and IBS-QOL
Carbone 2022, PMID 35483886 (DOMINO, NCT04270487) 459 primary-care IBS patients (41±15 years, 76% female, 70% Rome IV positive) randomised to 8 weeks of otilonium bromide 40 mg tds or a smartphone FODMAP-lowering diet application, followed 24 weeks Responders (IBS-SSS improvement ≥50) at 8 weeks: 71% (155/218) diet vs 61% (133/217) otilonium (p=0.03); in Rome IV-positive patients 77% (118/153) vs 62% (98/158), p=0.004. Advantage already present at 4 weeks (62% vs 51%, p=0.02) and response persisted through follow-up. Adherence 94% (199/212) diet vs 73% (148/202) drug, p<0.001. Predictors of response: female sex for diet (OR 2.08, p=0.04); PHQ15 for otilonium (OR 1.10, p=0.02)

The glucose-control trigger rate of 26% in the blinded reintroduction is the single most instructive number on this page: roughly a quarter of "FODMAP triggers" identified in open practice would be reproduced by an inert powder. That is a direct measurement of the nocebo component of dietary trigger identification — see placebo-response-and-trial-design.

Mechanism

A 2026 mechanistic trial in Rome IV IBS-D (NCT04542018; 42 patients, 34 responders) found that low FODMAP improved colonic barrier structure and function, mast cell number and mast cell mediator levels — but the magnitude of physiological change did not correlate with the magnitude of clinical response. Faecal supernatants from pre-diet patients caused barrier dysfunction when given intracolonically to wild-type mice; post-diet supernatants did not; the effect was abolished by lipopolysaccharide removal, TLR4 antagonism, tlr4−/− genotype, mast cell stabilisation or mast cell deficiency, and restored by reconstitution with wild-type but not tlr4−/− mast cells (Gao 2026, PMID 40749856). This is the most complete causal chain yet assembled for any IBS treatment: dietary FODMAPs → faecal LPS → TLR4 on colonic mast cells → barrier dysfunction. The non-correlation with symptom response is the caveat that keeps it from being the whole story.

Costs: the microbiome and the plate

  • Bacterial abundance and metabolites. Low FODMAP (3.05 g/day) versus typical Australian diet (23.7 g/day) for 21 days raised faecal pH (7.37 vs 7.16, p=0.001), reduced total bacterial abundance (9.63 vs 9.83 log10 copies/g, p<0.001) and altered composition; the Australian diet increased butyrate-producing Clostridium cluster XIVa (ratio 6.62, p<0.001) and Akkermansia muciniphila (19.3, p<0.001) (Halmos 2015, PMID 25016597).
  • Bifidobacteria and butyrate. In a 3-arm RCT (n=69), adequate symptom relief was higher with low FODMAP plus β-GOS than sham (16/24, 67% vs 7/23, 30%; OR 4.6, 1.3–15.6, p=0.015), but 1.4 g/day β-GOS did not prevent the bifidobacterial decline; Actinobacteria fell (1.9% and 1.8% vs 4.2% control) and faecal butyrate fell (387.3 and 346.0 vs 609.2). The authors' explicit conclusion: "strict long-term use should not be advised" (Wilson 2020, PMID 32433273).
  • Nutritional adequacy over the long run is better than feared. In 103 patients followed after dietitian-led low-FODMAP education, satisfactory symptom relief was 12% at baseline, 61% short-term and 57% long-term; 82% continued an "adapted FODMAP" diet (mean total FODMAP 20.6 ± 14.9 g/day) versus 18% returning to habitual intake (29.4 ± 22.9 g/day, p=0.039). Nutritional adequacy was not compromised in either group. The adapted group reported higher food cost (p<0.001) and more impact on social eating (p<0.01), with no difference in food-related quality of life or healthcare utilisation (O'Keeffe 2018, PMID 28707437).
  • Disordered eating. The relationship between adaptive food restriction in a GI condition and avoidant/restrictive eating pathology is contested rather than quantified: an expert review warns of both over- and under-pathologising eating behaviour in gastroenterology patients, and describes the "grey area" where restriction is an adaptive response to real physiological intolerance (Scarlata 2025, PMID 39681226; Atkins 2023, PMID 36782302). No cohort quantifying incident eating-disorder diagnoses after low-FODMAP instruction was retrievable in this session's searches (searched 2026-09-02). See also red-flags-and-safety-concerns and the anorexia nervosa condition in this knowledge base.

Delivery matters too: the low-FODMAP diet is dietitian-intensive, and the trials that produce the largest effects provide food (Halmos 2014, PMID 24076059) or dietitian-led education (O'Keeffe 2018, PMID 28707437) — neither of which is the routine primary-care experience.

Fibre

The 2026 synthesis (30 RCTs, 1,904 patients; searched to April 2026) is the current reference (Staudacher 2026, PMID 42600900):

Outcome Result
Clinical response (16 RCTs, n=1,293) 52% fibre vs 44% control; RR 1.21 (1.03–1.41), p=0.02, I²=38%
Psyllium specifically RR 1.53 (1.06–2.19)
Overall GI symptom scores (9 RCTs, n=497) SMD −0.25 (−0.67 to 0.17), p=0.25, I²=78% — not significant
β-galacto-oligosaccharides Improved symptoms on integrative scores
Commonest fibres studied Wheat bran (7 trials), psyllium (6), inulin-type fructans (5), β-GOS (2)

Subgroup analysis by IBS subtype was impossible for most outcomes — the authors call for future trials to report or recruit by subtype. Soluble fibre also features in the older network meta-analysis of soluble fibre, antispasmodics and neuromodulators (Black 2020, PMID 31859183), covered on antispasmodics-and-peppermint.

Gluten and wheat

Separating a gluten effect from a fructan effect has proved difficult, and self-reported non-coeliac gluten/wheat sensitivity is common in the population (Shiha 2025, PMID 41151790). A randomised single-blind gluten challenge in IBS has been performed (Saadati 2022, PMID 35322144), and a randomised double-blind placebo-controlled study addressed non-coeliac wheat sensitivity in refractory IBS and functional dyspepsia (Goyal 2025, PMID 41025571). Coeliac disease must be excluded before any gluten-restricted diet is started, since the serology and biopsy become unreliable off gluten — see differential-diagnosis-and-exclusion.

Lactose

Self-reported lactose intolerance in IBS has only moderate accuracy against hydrogen breath testing across six studies (845 participants) and generates frequent false positives; routine lactose-free diets and routine breath testing are not supported (Pop 2024, PMID 39255349). In the network meta-analysis, a lactose-reduced diet was one of six interventions studied in more than one trial but did not reach the top ranks (Cuffe 2025, PMID 40258374).

What the professional guidance says

The British Dietetic Association's systematic review and evidence-based practice guidelines for dietary management of IBS in adults (2016 update) remain the reference dietetic document (McKenzie 2016, PMID 27272325), with a companion review for probiotics (McKenzie 2016, PMID 27265510). Full guideline comparison is on guidelines.

Open questions

  • Is a starch- and sucrose-reduced diet genuinely the best diet for IBS, or an artefact of two small trials (Cuffe 2025, PMID 40258374)?
  • Does the CARIBS result — diet beating optimised drugs — replicate outside a single specialist Swedish centre with dietitian support (Nybacka 2024, PMID 38643782)?
  • Why does low FODMAP beat habitual diet but not traditional dietary advice (Böhn 2015, PMID 26255043; Cuffe 2025, PMID 40258374)? Is the active ingredient FODMAP restriction or structured dietary attention?
  • What is the true nocebo fraction of "identified food triggers"? The blinded reintroduction control (glucose) triggered recurrence in 26% of exposures (Van den Houte 2024, PMID 38401741).
  • What are the long-term consequences of reduced Bifidobacterium, Actinobacteria and butyrate (Wilson 2020, PMID 32433273; Halmos 2015, PMID 25016597)? A 12-month randomized follow-up has now measured symptoms and microbiome composition after a personalised diet versus low FODMAP; the low-FODMAP benefit regressed by 12 months and no significant between-group beta-diversity difference remained at that point (Tunali 2026, PMID 42623122). It did not test the clinical consequences of the specific Bifidobacterium, Actinobacteria or butyrate reductions reported in the earlier trials, which remains the dated evidence gap as of 2026-09-02.
  • Does dietary restriction cause disordered eating in IBS? The question is framed but not quantified (Scarlata 2025, PMID 39681226).
  • Why does barrier and mast-cell improvement not correlate with symptom improvement on the same diet (Gao 2026, PMID 40749856)?
  • Which IBS subtype benefits most from which fibre? Not answerable from current trials (Staudacher 2026, PMID 42600900).

References

  1. Nybacka S, et al. A low FODMAP diet plus traditional dietary advice versus a low-carbohydrate diet versus pharmacological treatment in irritable bowel syndrome (CARIBS). Lancet Gastroenterol Hepatol. 2024;9(6):507-520. PMID 38643782
  2. Cuffe MS, et al. Efficacy of dietary interventions in irritable bowel syndrome: a systematic review and network meta-analysis. Lancet Gastroenterol Hepatol. 2025;10(6):520-536. PMID 40258374
  3. Halmos EP, Power VA, Shepherd SJ, Gibson PR, Muir JG. A diet low in FODMAPs reduces symptoms of irritable bowel syndrome. Gastroenterology. 2014;146(1):67-75.e5. PMID 24076059
  4. Böhn L, et al. Diet low in FODMAPs reduces symptoms of irritable bowel syndrome as well as traditional dietary advice: a randomized controlled trial. Gastroenterology. 2015;149(6):1399-1407.e2. PMID 26255043
  5. Altobelli E, Del Negro V, Angeletti PM, Latella G. Low-FODMAP Diet Improves Irritable Bowel Syndrome Symptoms: A Meta-Analysis. Nutrients. 2017;9(9):940. PMID 28846594
  6. Van den Houte K, et al. Efficacy and Findings of a Blinded Randomized Reintroduction Phase for the Low FODMAP Diet in Irritable Bowel Syndrome. Gastroenterology. 2024;167(2):333-342. PMID 38401741
  7. Carbone F, et al. Diet or medication in primary care patients with IBS: the DOMINO study. Gut. 2022;71(11):2226-2232. PMID 35483886
  8. Di Rosa C, et al. DOMINO trial post hoc analysis: evaluation of the diet effects on symptoms in IBS subtypes. Therap Adv Gastroenterol. 2024;17:17562848241255296. PMID 39086991
  9. Tack C, et al. A study on the utility of inflammatory biomarkers in primary care irritable bowel syndrome: a sub analysis of the DOMINO randomized trial. BMC Gastroenterol. 2025;26:69. PMID 41436948
  10. Routhiaux K, et al. Involvement of Eosinophil-Driven Intestinal Immune Activation in Different Irritable Bowel Syndrome Subtypes and in the Response to a FODMAP Lowering Diet: A Post Hoc Analysis of the Randomized Controlled DOMINO Trial. Gastroenterology. 2026;170(4):818-820. PMID 41670574
  11. Tunali V, et al. A Multicenter Randomized Controlled Trial of Microbiome-Based Artificial Intelligence-Assisted Personalized Diet vs Low-FODMAP Diet. Am J Gastroenterol. 2024;119(9):1901-1912. PMID 38717025
  12. Gao J, et al. Low FODMAP Diet Improves Colonic Barrier Function and Mast Cell Activation in Patients With Diarrhea-Predominant Irritable Bowel Syndrome: A Mechanistic Trial. Gastroenterology. 2026;170(1):132-147. PMID 40749856
  13. Halmos EP, Christophersen CT, Bird AR, Shepherd SJ, Gibson PR, Muir JG. Diets that differ in their FODMAP content alter the colonic luminal microenvironment. Gut. 2015;64(1):93-100. PMID 25016597
  14. Wilson B, et al. β-Galactooligosaccharide in Conjunction With Low FODMAP Diet Improves Irritable Bowel Syndrome Symptoms but Reduces Fecal Bifidobacteria. Am J Gastroenterol. 2020;115(6):906-915. PMID 32433273
  15. O'Keeffe M, et al. Long-term impact of the low-FODMAP diet on gastrointestinal symptoms, dietary intake, patient acceptability, and healthcare utilization in irritable bowel syndrome. Neurogastroenterol Motil. 2018;30(1). PMID 28707437
  16. Staudacher HM, Rucco V, Mancell S, Dimidi E, Whelan K. Fiber Supplementation in Irritable Bowel Syndrome: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Gastroenterology. 2026 Aug 14 (online ahead of print). PMID 42600900
  17. Scarlata K, et al. A Call to Action: Unraveling the Nuance of Adapted Eating Behaviors in Individuals With Gastrointestinal Conditions. Clin Gastroenterol Hepatol. 2025;23(6):893-901.e2. PMID 39681226
  18. Atkins M, et al. Assessment and management of disorders of gut-brain interaction in patients with eating disorders. J Eat Disord. 2023;11:20. PMID 36782302
  19. Pop A, et al. Self-Perceived Lactose Intolerance Versus Confirmed Lactose Intolerance in Irritable Bowel Syndrome: A Systematic Review. J Gastrointestin Liver Dis. 2024 Sep 9 (online ahead of print). PMID 39255349
  20. Shiha MG, et al. Global prevalence of self-reported non-coeliac gluten and wheat sensitivity: a systematic review and meta-analysis. Gut. 2025 Oct 28 (online ahead of print). PMID 41151790
  21. Saadati S, et al. Effects of a gluten challenge in patients with irritable bowel syndrome: a randomized single-blind controlled clinical trial. Sci Rep. 2022;12:4960. PMID 35322144
  22. Goyal O, et al. Prevalence and predictors of nonceliac wheat sensitivity in refractory irritable bowel syndrome and functional dyspepsia: results from a randomized double-blind placebo-controlled study. Eur J Gastroenterol Hepatol. 2025;37(11):1238-1248. PMID 41025571
  23. McKenzie YA, et al. British Dietetic Association systematic review and evidence-based practice guidelines for the dietary management of irritable bowel syndrome in adults (2016 update). J Hum Nutr Diet. 2016;29(5):549-75. PMID 27272325
  24. McKenzie YA, et al. British Dietetic Association systematic review of systematic reviews and evidence-based practice guidelines for the use of probiotics in the management of irritable bowel syndrome in adults (2016 update). J Hum Nutr Diet. 2016;29(5):576-92. PMID 27265510
  25. Black CJ, et al. Efficacy of soluble fibre, antispasmodic drugs, and gut-brain neuromodulators in irritable bowel syndrome: a systematic review and network meta-analysis. Lancet Gastroenterol Hepatol. 2020;5(2):117-131. PMID 31859183
  26. Tunali V, et al. Long-term microbiome and clinical effects of a microbiome-guided personalized diet versus low-FODMAP diet in irritable bowel syndrome: A 12-month follow-up randomized controlled trial. Gut Microbes. 2026;18(1):2719125. PMID 42623122